Reprinted with permission from

PACING AND CLINICAL ELECTROPHYSIOLOGY, Volume 17, NO. 8, August 1994

Copyright O 1994 by Futura Publishing Company, Inc., Armonk, NY 10504-0418

Digital Signal Processing Chip Implementation

for Detection and Analysis of Intracardiac
Electrograms
CHIH-MING JAMES CHIANG, JANICE M. FNKINS,
and LORENZO A. DICARLO
From the Department of Electrical Engineering and Computer Science, College of Engineering,
University of Michigan, and The Michigan Heart and Vascular Institute and Cardiac
Electrophysiology Laboratory, St. Joseph Mercy Hospital, Ann Arbor, Michigan

CHIANG, C.-M.J., ET AL.: Digital Signal Processing Chip Implementation for Detection and Analysis of
Intracardiac Electrograms. The adoption of digital signal processing (DSP) microchips for detection and
analysis of electrocardiographic signals offers a means for increased computational speed and the oppor-
tunity for design of customized architecture to address real-time requirements. A system using the Moto-
rola 56001 DSP chip has been designed to realize cycle-by-cycle detection (triggering) and waveform
analysis using a time-domain template matching technique, correlation waveform analysis (CWA). The
system digitally samples an electrocardiographic signal at 1000 Hz, incorporates an adaptive trigger for
detection of cardiac events, and classifies each waveform as normal or abnormal. Ten paired sets of
single-chamber bipolar intracardiac electrograms (1-500 Hz) were processed with each pair containing
a sinus rhythm (SR) passage and a corresponding arrhythmia segment from the same patient. Four of ten
paired sets contained in traatrial electrograms that exhibited retrograde atrial conduction during ventricu-
lar pacing; the remaining six paired sets of intraventricular electrograms consisted of either ventricular
tachycardia (4) or paced ventricular rhythm (2). Of 2,978 depolarizations in the test set, the adaptive
trigger failed to detect 6 (99.8% detection sensitivity) and had 11false triggers (90.6% specificity). Using
patient dependent thresholds for CWA to classify waveforms, the program correctly identified 1,175 of
1,I 97 (98.2% specificity) sin us rhythm depolarizations and 1,771 of 1,781 (99.4 % sensitivity) abnormal
depolarizations. From the results, the algorithm appears to hold potential for applications such as real-
time monitoring of electrophysiology studies or detection and classification of tachycardias in implantable
antitachycardia devices. (PACE 1994; 17:1373-1379)

arrhythmia, intracardiac electrogram, antitachycardia device, implantable defibrillator

Introduction amplitude, area under the curve) were advanced

for diagnostic electrocardiograms (ECGs).' As
Mor~hological Of cardiac waveforms computer techniques became more
has long been a in
better methods were used, particularly in the clas-
diOgra~h~. measures sification of and abnormal depolarizations
for such applications as mal-time coronary care
monitorine" and fast-time scanning of ambulatory
This work was partially supported by National Science Foun- (Halter) monitoring. The method choice in these
dation Grants BCS-8909042 and EID-9023514.
applications converged in the late 1970s to a point-
Address for reprints: Chih-ming J Chiang. Ph.D.. Telectmnics
Pacing Systems, 7400 S. Tucson Way, Englewood, CO 80112. bypoint comparison of each digitized waveform
Fax: (303) 799-2213. with a previously extracted template of a normal
Received June 25, 1993; revision January 10, 1994; accepted d e p ~ l ~ i z a t i oThis
n . ~ technique
~~ used the classic
February 2, 1994. statistical correlation coefficient? and later be-

PACE, Vol. 17 August 1994 1373

 CHIANG, ET AL.

came known as the correlation waveform analysis Adjustable Threshold Trigger

(CWA].5-7CWA is a robust measure; it is imper- An accurate triggering mechanism is essen-
vious to the DC offset of the ECG and to amplitude tial for rate and morphological analysis of intra-
variations of the waveforms. More recently, CWA cardiac electrograms. A previously developed
has been applied to morphological analysis of in- multistage software depolarization detector using
tracardiac electrograms for recognition of abnor- adjustable thresholds has been shown to be ro-
mal activation recorded by catheter electrode^.^ It bust in detecting intracardiac electrogramsl' and
has been shown to be equally robust in this setting, is the basis for the trigger implemented on the
given that appropriate bandwidths of the recorded DSP chip.
signal be o b ~ e r v e d Due
. ~ to the heavy computa- ' The scheme contains a second order IIR But-
tional demands of CWA, however, its only real- terworth band-pass filter that suppresses unde-
time application has been limited to coronary care sired frequency components and eliminates base-
monitoring systems. line shift. The trigger has an adaptable threshold
One application that would benefit from in-
with exponential time decay to allow for variabil-
corporation of morphological methods are anti- ity in waveform amplitude. A blanking period is
tachycardia devices such as pacing and nonpacing imposed to prevent multiple event detection on a
implantable cardioverter defibrillators. These de-
single depolarization.
vices are designed to terminate potentially life- Figure 1 shows a comprehensive diagram of
threatening tachyarrhythmias such as ventricular the trigger. The first stage of the filter is a digital
tachycardia (VT) and ventricular fibrillation [VF)
band-pass IIR filter H(z) derived from the bilinear
and have been shown to perform well in prevent-
ing sudden cardiac death.' However, these-devices
utilize simple detection schemes based upon rate
and yield a high incidence of false delivery of ther-
apy due to the one-channel nature of detection,
which frequently confuses supraventricular ar-
rhythmias with VT and VF. It has been suggested Bandpass
2060 Hz
that morphological analysis is the key to alleviate
this p r ~ b l e m .To
~ ?date,
~ however, the only real-
time programs using CWA for intracardiac elec-
trogram waveform a n a l y s i ~are ~ ~implemented
~~
on the PC-based 386 computer (Intel P700, Intel
Corp., Beaverton, OR, USA) or coronary care moni-
toring machines.
The goal of this study was to demonstrate the
feasability of implementing CWA on a digital sig-
nal processing (DSP) chip (Motorola 56001 DSP,
Motorola Communications & Electronics, Inc.,
Schaumburg, IL, USA). This would represent a
first step towards specialized architecture incor-
porating CWA that would be utilized in future gen-
eration implantable antitachycardia devices.
Trigger
a d
Blank tor
next 150mS
Methods
The DSP program contains main Figure 1. The block diagram for the trigger. x = input,
subsections, the first a depolarization detection w
= outputfromthe bandpassfilter, = the magnitude
scheme described in the "Adjustable Threshold ,fy, = fraction to multiply with W, d = fraction to
Trigger" subsection, and the second the CWA de- multiply with previous vvalue, v = value from Equation
scribed in the CWA subsection. 2, Max = maximum.

1374 August 1994 PACE,Vol. 17

 DSP CHIP IMPLEMENTATION OF CORRELATION WAVEFORM ANALYSIS
transform of a second order analog filter H(s) with
s = (1 - z- l)I(l + zF1)ll:
and ao, a,, a, are derived as follow^'^:
where WL, W H are:
with f~ the lower cut-off frequency, fH the higher
cut-off frequency, and f, the sampling frequency
of the system. The resulting difference equation is:
Yi = 00 (xi - xi-,) - alyj-1 - a~yi-, (I)
where yi and xi are the ith filter output and input,
respectively.
The second stage is a threshold comparison of
the magnitude of the filtered signal with an adjust-
able threshold. The threshold v is the maximum
of two components, the first a fraction (r) of the
rectified signal and the second a decay factor (d
with d < 1) multiplied with the previous thresh-
old. The update function for the threshold is:
The two factors, r a n d d, control the sensitivity of
the trigger to allow for detection of waveforms
with varying amplitudes. Making r small and d
small will increase sensitivity of detection at the
expense of increased false triggering probability.
The third stage is a blanking interval that prevents
multiple triggering during a single depolarization.
For this study, values were set as follows: fL = 20
PACE, Vol. 1 7 August 1994
Hz, fH = 60 Hz, d = 2-1'1000,and r = 0.5 based
upon empirical results.ll
CWA
The CWA performmce measure p, indepen-
dent of amplitude fluatuations and baseline
changes, yields an output between - 1 and + 1
where + 1 indicates a perfect match. Mathemati-
cally, p is defined as1':
where N = the number of points in the template;
ti = the template points; si = the signal points to
be processed; t = the template average; and = s
the signal average.
To further reduce computation, we eliminate
the square root calculation, and rename the met-
ric C3 = sign(p)$ where sign is f 1 depending
on the sign of p. If normalization of the template
points has been calculated (such as by an external
programmer) in advance such that i = 0 and
y t i = 1,then the on-line calculation of newly
introduced depolarizationg becomes12:
[ZjIy (ti)(si))2
(4)
This represents a streamlined version of CWA
(using an N point template), which requires 2N +
2 multiplications, 1 division, and 3N + 1 addi-
tions per waveform. This is the equation that is
implemented on the DSP chip.
Motorola 56001 System
The Motorola 56001 DSP chip was designed to
perform fast multiply and accumulate operations
appropriate for DSP procedures, such as filtering
and Fast Fourier Transforms (FFT).13.14The chip
is based on a modified Harvwd architecture13with
a 512 word dynamic random access program mem-
ory (DRAM) and two on-chip X and Y data read-
only-memory (ROM)as well as data random access
memory (RAM) (256 in X and 256 in Y for both
RAM and ROM).13.14It has a 24-bit data register
 CHIANG, ET AL.
and two accumulators of 56-bit precision. It exe-
cutes in fixed point arithmetic with a computa-
tional speed of 10.25 million instructions per sec-
ond (MIPS).13.14The system contains a 16-bit ana-
log to digital (AID) converter with a maximum
sampling rate of 44 ~ H z . ' ~ , ' ~
The implemented DSP program contains three
main components, the analog to digital interface,
the triggering section, and the CWA section. Essen-
tially, the algorithm remains in a wait state until
an interrupt occurs indicating that data have been
received from the AID. Upon this event, the pro-
gram stores the data and enters the triggering sec-
tion. If a depolarization occurs, CWA is activated.
CWA is implemented using Equation 4 to reduce
computational load. Template values are normal-
ized prior to utilization and a 64 point CWA (64
msec at a 1,000Hz sampling rate) was chosen since
most intracardiac waveforms fall within this dura-
tion and since 64 is a power of 2 (z6). To align the
template with the waveform being analyzed, the
window centered around the trigger for CWA cal-
culation was shifted by n points, with n varying
from - 10 to + 10, creating 2 1 separate windows.
CWA was performed for each of the windows and
the maximum value of the 2 1 taken as the true
CWA performance measure. If this measure was
greater than a patient dependent threshold of 0.9
or 0.81 (since signal amplitudes and variabilities
were patient dependent, the threshold were cho-
sen individually for each pair of passages), the DSP
chip output a positive pulse for a normal depolari-
zation, otherwise it gave a negative pulse indicat-
ing an abnormal waveform.
Materials
Ten pairs of single channel bipolar intracar-
diac electrograms (1-500 Hz) from seven patients
were processed with each pair consisting of one
sinus rhythm (normal) passage and one corre-
sponding arrhythmia segment. Four of the pairs
were intraatrial electrograms containing retro-
grade conduction during ventricular pacing. The
other six pairs were intraventricular electrograms
containing ventricular tachycardia (4) or ventricu-
lar pacing (2).
Data used for analysis were recorded from 6-
French quadrapolar electrode catheters (USCI Di-
vision, C.R. Bard Inc., Billerica, MA, USA) with
an interelectrode distance of 1cm. The recordings
1376 August 1994
were made in the high right atrium and right ven-
tricular apex on FM magnetic tape (Hewlett-Pack-
ard Model 3968A, San Dieglo, CA, USA) at 3:
inches per second during prov~cativeelectrophys-
iology studies with patients lying supine. For pro-
cessing, data were digitized through the DSP AID
at 1,000 Hz and subjected to analysis on the DSP
chip.
Results
Of the 2,978 total depolaL.izationsfrom the 10
pairs studied, the adaptive trigger missed only 6
(99.8% sensitivity of depolmization detection)
while giving 11false-positives (99.6% specificity).
Results are shown in Table I. For the performance
of CWA with patient dependent thresholds and
shifting window calculations (see Motorola 56001
System), the program correctly identified 1,171 of
1,197 (97.8% specificity) sinus rhythm (normal)
depolarizations and 1,771 of 1,781 (99.4% sensi-
tivity) abnormal depolarizations. The overall re-
sult was 2,943 of 2,978 (98.8%) (Table 11).
Figure 2 contains intraventricular electrogram
results from patient 4. The lleft side of the figure
depicts a sinus rhythm passge along with the di-
agnostic signals and the right side has a VT seg-
ment with its correspondin$ program diagnosis.
As can be seen from the sinus rhythm passage, trig-
ger and CWA sections functioned properly with
the periodic positive pulse outputs indicating nor-
mal waveform diagnosis. For the VT passage, the
trigger works equally well and the negative pulse
reflects abnormal waveform diagnoses.
Figure 3 contains intraatrial electrogram re-
sults from a sinus rhythm passage followed by a
ventricular pacing section (yielding retrograde
atrial activation) with its corresponding diagnostic
pulses. There are no false trliggers and all normal
depolarizations are indicated in the event marker
by positive pulses and abnormal ones by negative
pulses. The black blotches represent discontinu-
ities during the taping process and their corre-
sponding results are disregwded when tabulating
system performance.
Figure 4 contains an intraatrial electrogram
passage that revealed triggering failures. The be-
ginning of the trace shows d n u s rhythm followed
by ventricular pacing. Note the false-positive trig-
gering in the pacing segment due to a wandering
PACE, Vol. 17
 DSP CHIP IMPLEMENTATION OF CORRELATION WAVEFORM ANALYSIS

Table I.
Result for Real-Time lm~lementationof the Triaaer

Sinus Rhythm Arrhythmia

Patient Channel Num FP FN Num FP FN

1
2
3
4
5
6
7
8
9
10
Total
Grand Total
- - -- -- -

Num = number; FP = false-positive; FN = false-negative.

baseline. CWA also failed in one depolarization
during sinus rhythm when a normal waveform
was declared abnormal.
Discussion and Conclusion
As shown previously, the DSP system per-
formed well in waveform detection and CWA on
intracardiac electrograms, The trigger obtained an
accuracy of 99.4%. The few cases of missed trig-
gering (612,968)resulted from a sequence in which
a large amplitude waveform was followed by one
of small amplitude, such that the threshold re-
mained too high for the second depolarization to
be detected. ~aise-positivetriggers occurred due to
wandering baseline coupled with small amplitude
-

Table 11.
Result for Real-Time lm~lementationof CWA

Normal Abnormal

Patient Channel Num Wrong Num Wrong

1 V 170 13 105 0
2 V 84 2 89 0
3 V 191 1 271 6
4 V 137 0 214 0
5 A 146 7 249 0
6 A 70 0 229 0
7 A 59 0 143 0
8 V 50 1 134 0
9 A 164 0 156 4
10 v 126 -
2 191 -0
Total 1,197 26 1.781 -
10
Grand Total 2,978 36

CWA = correlation waveform analysis; Num = number.

PACE, Vol. 1 7 August 1994

 CHIANG,ET AL.

Ventricular Electrogram

SINUS RHYTHM VENTRICULAR TACHYCARDIA

DSP Output
Figure 2. Patient 4 intraventricular electrograms are shown. The left side of the figure contains
a sinus rhythm passage with accompanying diagnostic signals and the right side has a ventricular
tachycardia passage along with the diagnosis.

signals. The small incidence of these events does
not warrant modification of the triggering mecha-
nism. A previous study examining the perfor-
mance of this depolarization detection scheme
showed that the trigger performed well." The
Medtronic PCD (Medtronic, Inc., Minneapolis,
MN,USA) detection scheme is also similar to the
trigger presented in this study.
For CWA measures, the program was able to
distinguish between normal and abnormal wave-
forms with at least 98% accuracy using patient de-
pendent thresholds. Most errors were normal
waveformsclassified as abnormal due to low CWA
values and these were mainly due to inaccuracy
of the trigger location. The 21-window CWA calcu-
lation alleviates most of the problem of inexact-
ness of depolarization detection.
In conclusion, a real-time program has been

Atrial Electrogram

SINUS RHYTHM VENTRICULAR PACING

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 l l l l l l l l l l l l ~ l l l l

DSP Output
Figure 3. Intmatrial electrogram results. Sinus rhythm is followed by pacing.

August 1994 PACE, Vol. 17

 DSP CHIP IMPLEMENTATION OF CORRELATION WAVEFORM ANALYSIS

Atrial Electrogram

VENTRICULAR PACING

/ ' I l l 1 I L L l l l l l
1111I I ~ I I I I I I I . I ~ ~ ~ I I I I I ~ I
AA AAAAA
DSP Output False Positives (FP)
Figure 4. An intraatrial passage with sinus rhythm followed by ventricular pacing. This shows
how the wandering baseline corrupts the trigger peqormance with false-positive (FP) detection
of waveforms. Notice also the incorrect digital signal processing (DSP) output of negative pulse
on sinus rhythm passage due to low correlation value.

implemented on the Motorola 56001 DSP chip that miniaturization that would make incorporation of
detects cardiac activation and performs CWA on CWA possible in real-time applications such as
intracardiac electrograms. The program performs next-generation implantable antitachycardia de-
well and represents an important first step towards vices.

References
Jenkins J. Automated electrocardiography and ar-
rhythmia monitoring. Prog Cardiovasc Dis 1983;
25:367-408.
Arzbaecher R, Biancalana P, Stibolt T, et al. Com-
puter technique for detection of cardiac arrhyth-
mias. (abstract) J Assoc Advance Med Instrument
1971; 5:104.
Feldman C, Arnazeen P, Klein M, et al. Computer
detection of ventricular ectopic beats. Comput Bio-
med Res 1971; 4:666-674.
Kennedy JB, Neville AM. Basic statistical methods
for engineers and scientists. Harpers & Row, Pub-
lishers, New York, NY, 1986, pp. 410-411.
Lin D, DiCarlo L, Jenkins J. Identification of ven-
tricular tachycardia using intracavitary ventricular
electrograms: Analysis of time and frequency do-
main patterns. PACE 1988; 11:1592-1605.
DiCarlo L, Throne R, Jenkins J. A time-domain
analysis of intracardiac electrograms for arrhyth-
mia detection. PACE 1991; 14:329-336.
Throne R, Jenkins J, Winston S, et al. Discrimina-
tion of retrograde from anterograde atrial activa-
tion using intracardiac electrogram waveform
analysis. PACE 1989; 12:1622-1630.
Cohen T, Chien W, Lurie K, et al. Implantable car-
dioverter defibrillator proarrhythmia: Case report
and review of the literature. PACE 1991; 14:
1326-1329.
9. Chiang CJ, Jenkins JM, DiCarlo LA. An innovative
two-channel rate and morphology method for
complex arrhythmia diagnosis in implantable de-
vices. 14th Annual International Conference of the
IEEE Engineering in Medicine and Biology Soci-
ety, 1992;496-497.
10. Chiang CJ, Jenkins JM, DiCarlo LA. Real-time ar-
rhythmia identification from automated analysis
of intraatrial and intraventricular electrograms.
PACE 1993; 16(1):223-227.
11. McDonald R, Jenkins J, Arzbaecher R, et al. A soft-
ware trigger for intracardiac waveform detection
with automatic threshold adjustment. IEEE Com-
puters in Cardiology, IEEE Press, New York, NY,
1989, pp. 167-170.
12. Throne RD.Analysis of template matching meth-
ods for intracardiac electrogram analysis. Ph.D.
Thesis, University of Michigan, 1990.
13. Lee E. Programmable DSP Architectures: Part I.
IEEE ASSP Magazine 1988; 5(4):4-19.
14. Lee E. Programmable DSP Architectures: Part 11.
IEEE ASSP Magazine 1989; 6(1):4-14.

PACE, Vol. 17 August 1994 1379